The Nondestructive Inspection Research Of Organic Phosphorus Pesticide Residues In Tuber Vegetable Based On Near Infrared Spectroscopy

Pesticide residues threatened is the most important factors to the safety of agricultural products. However, the traditional pesticide residue detection methods have a lot of disadvantages, such as destructive, have a long test cycle,.high cost, cause environmental pollution and lacks for real time. Near infrared spectroscopy, as a kind of fast and nondestructive "green detection" technology, is more and more concerned in agricultural products detection in recent years. As a kind of tuber vegetable, ginger is essential condiment for everyone in daily lives, at the same time it is a kind of important medicinal herbs, and in some areas of China it’s also one type of important export vegetables. However, in order to get higher economic interests, during the growth of gingers,growers always use a large number of pesticides, especially organophosphorus pesticide to control cut production. To evaluate the potential of nondestructive testing of organophosphorus pesticide residues in tubers vegetable’s, in this paper, tubers vegetables ginger as the research object, use dichlorvos and methamidophos two kinds of organophosphorus pesticides as the representative, use of Near Infrared Spectroscopy Nondestructive Testing dichlorvos and methamidophos content in the ginger, The main content of this paper are as follows:1、Use of near infrared spectral technology detection the dichlorvos pesticide residues in gingers. Use the near infrared spectrometer acquisition band380～1800nm sample spectra, First derivative (FD), second derivative (SD), standard normal variate (SNV) and multiplicative scattering correction (MSC) were for the acquisition of Vis-NIR spectra. Remove the noise more of the band380～529nm, In the band530～1800nm use partial least squares (PLS) and the least square support vector machine (LS-SVM) were established in the content of gingers difference of near infrared spectral quantitative analysis model. Through to the prediction of the set forecast, four kinds of pretreatment method used in the first derivative pretreatment method, using the least square support vector machine (SVM) established in the content of gingers dichlorvos quantitative analysis model optimal, The validation group of correlation coefficient and RMSECV were0.8879and17731respectively. The forecasting group of correlation coefficient and RMSEP were0.8862and2.2314respectively. The results suggest that it is feasible to develop near infrared spectral technology as an on-line sensor for the dichlorvos pesticide residues in gingers.2、Based on the nondestructive testing of the methamidophos residues in the ginger by near infrared spectral technology nondestructive testing the methamidophos residues in the ginger. After Spectrum data for standard orthogonal transform (SNV) after pretreatment, use full spectrum PLS, iPLS and siPLS are established quantitative analysis model, compare with the three methods to establish the model of the results are found, using Synergy interval partial least square (siPLS) establish gingers quantitative analysis of methamidophos content model optimal, the spectrum be divided into21areas, adopt main composition of number is8, Joint sub-intervals were6,7,8and13. The validation group of correlation coefficient and RMSECV were0.8436and1.5062respectively. The forecasting group of correlation coefficient and RMSEP were0.8146and1.7295respectively. The results of the study show that both of iPLS and siPLS can reduce the amount of calculation model and improve the efficiency of the precision of the model, especially siPLS effect better. The research shows that it realizes the nondestructive testing by using of near infrared spectral technology in combination with appropriate spectrum data analysis method of gingers in pesticide methamidophos.